►
From YouTube: 2020-11-17-Package Maintenance Team meeting
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
we'll
follow
the
agenda
as
outlined
in
the
issue
in
the
repo
which
is
issue
number
430.
Before
we
get
started,
does
anybody
have
any
announcements
they'd
like
to
share.
A
B
B
Oh
okay,
so
yeah
I
posted
this.
I
was
yeah.
Basically,
if,
if
you
missed
the
issue,
I
was
kind
of
poking
around
and
trying
to
figure
out
well
what
what
should
I
work
on
here?
Because
I
want
to
get
more
involved
here,
and
so
I
found
some
things
and
I
wasn't
they
did
they
didn't
look
like
they
were
up
to
date.
Essentially
so
there's
this
roadmap
github
project
and
there's
some
stuff.
You
know
it's
one
of
those
kanban
github
boards
and
there's
some
stuff
in
there
and
it
says
in
progress.
B
B
Right,
so
maybe
this
this
project,
if
it's
not
in
use
you
know,
maybe
we
should
just
get
rid
of
it
or
something
but
anyway.
So
in
this
issue,
I
I
listed
the
ones
that
were
marked
as
in
progress,
and
perhaps
that
doesn't
really
matter
if
we're,
if
we're
not
really
using
this
thing,
but
the
other
one
was.
B
I
was
looking
through
the
issues
and
a
bunch
of
these
were
marked
good
first
issue
and
the
the
vast
majority
of
these
seem
to
be
requests
for
draft
guidance
documents,
and
I
I
it's
there
was,
you
know
a
handful
of
them.
I'm
not
sure
you
know
what
is
going
on
with
those
if,
if
they're
still
needed,
if
those
really
are
like
a
good
first
issue
at
all,
because
it
would
seem
to
me
like
that.
Those
like
a
guidance
document
is
something
like
that.
We
want
to
really
collaborate
on.
C
B
As
a
team,
instead
of
I
don't
know,
but
maybe
somebody
can
just
come
along
and
send
a
draft,
but
I
I
feel,
like
that's
probably
unlikely
so
yeah
there
was
there-
was
that-
and
I
guess
a
a
few
more
questions
in
this
issue,
but
I
don't
know
what
I
guess
we
can
start
with
the
the
good
first
issue
thing:
if
we're,
if
we're
we've
determined
we're
not
really
using
that
github
project,
so
the
good
first
issue
stuff.
So
maybe
these
guidance
documents.
Where
are
we
at
with
them.
C
So
so
I
was
just
about
to
start
off
the
integration
testing
one
because
I
made
a
promise
to
link
it
back
to
the
unit
testing
one
yeah
I
remember,
and
so
we
could
actually
work
on
that
I've
been
very
busy
up
to
now
with
the
other.
You
know
the
work
work
side
of
things,
but
I
should
be
getting
some
time
coming
up
now
and
I'll
be
happy
to
work
with
you,
christopher.
B
Okay,
so
there's,
let's
see
it
looks
like
you're
assigned
to
that
one
and
then
this
other
one
is
this
testing
guidelines.
C
Yes,
yeah
the
testing
and
they
can
they
cut.
They
will
obviously
refer
to
each
other,
and
I
made
some
commitments
that
I
would
refer
back
to
some
certain
points
in
some
of
the
meetings
and
and
I'll
be
happy
to
work
on
that
with
you.
As
you
said,
it's
a
collaborative
one
and
if
you're
the
christopher,
I
think
you
are.
You
should
be
good
at
testing.
B
A
Basically,
drafts
and
the
ones
that
are
just
in
the
docs
directory
which
are
promoted,
so
I
think
we
might
have
opened
some
of
those
to
promote
the
current
drafts
in
addition
to
writing
them
from
scratch.
So
it
kind
of
be
like
for
each
one
of
those
it's
like.
Is
there
one
that's
already
in
a
draft?
If,
yes,
you
know
review
it
to
help,
get
it
promoted
to
non-draft.
A
B
D
So
there's
the
the
ones
marked
as
drafts
are
the
maybe
a
bit
misleading.
It's
not
a
the
call
to
action,
isn't
to
draft
a
first
version
of
them.
The
the
call
to
action
is
what
michael
relayed,
which
is
their
in-draft.
A
lot
of
them
are
pretty
complete,
and
so
what
we're
hoping
is?
Could
someone
just
come
pick
it
up,
pull
it
out
a
giraffe.
You
know
give
that
review
open
a
pr.
So
then
that
could
pull
the
rest
of
us.
D
D
Yeah
and
then
there's,
I
think,
there's
some
I
think
there
is.
There
are
a
couple
call
to
actions
for
just
brand
new
documentation.
I
think
we
did
an
audit
and
found
a
couple
topics
that
don't
exist,
that
we
would
like,
as
top
level,
so
those
those
aren't
the
draft
promotion,
so
it
yeah,
so
it
depends,
but
we're
yeah,
hoping
that
just
moving
a
file
and
adding
it
to
a
table
of
contents
and
giving
it
a
once-over
would
be
maybe
some
potentially
enticing
work.
But
you
know
it's
that's
docs
too
so
gonna.
A
Right
so
actually
yeah,
so
those
ones
most
a
number
of
those.
It
would
actually
be
good
to
it's
been
a
while
and
it
was
like
if
we,
if
we
were
they're,
basically
done,
we
should
get
promoted.
So
that
would
be
a
good
thing
for
some
for
you
to
do
chris
to
just
actually
create
the
prs
to
move
them
to
non-draft.
D
D
Yeah,
I'm
sure
they
could
benefit
from
you
know.
D
B
D
If
you
look
at
my
comments
on
428,
it
kind
of
captures
that
so
it's
basically
two
step
well,
three
steps
proofread
actually
do
like
a
move
from
the
draft
folder
to
the
docs
folder
and
then
there's
a
table
of
contents
in
the
of
a
readme
file
in
the
docs
folder,
and
so
we
would
just
want
to
mark
that
file
as
no
longer
being
a
draft
anymore.
B
D
B
All
right-
and
so
this
kind
of
ties
into
another
question
I
had
then
on
in
that
428,
which
was
like
okay,
so
we
have
these
documents
they're
in
this
docs
folder,
no
one's
going
to
find
them
there
right.
So
how
do
we
get
these
in
front
of
maintainers.
D
D
So
a
little
bit
they
are.
If
you
go
to
the
readme,
the
second
section
is
called
for
maintainers
and
right
and
it
links
to
them
there.
So
that's
one
spot
and
then
hopefully
you
know
we're
hoping
that,
as
you
know,
more
tools
and
stuff
come
out
of
this
community.
They
could
also
go
under
that
maintainers
section.
So
that's
that's
the
first
one
and
then
you
know
the
other.
The
table
of
contents
was
added
because
we
didn't
at
the
time
didn't
have
like
a
single
listing
of
all
of
them.
D
A
A
At
the
point
you
know
we
we
we
built
them
up
over
time.
They
were
in
draft
there's
enough
now,
especially
if
we
get
those
other
ones
promoted
that
it's
like
you
know,
we
need
to
come
up
with
our
strategy
for
making
them
visible,
you
know
and
promoting
them
every
like.
I
don't
you
know,
I
mean
sharing
them
on.
D
Well,
I
think
it
goes
to
another
point
that
christopher
brought
up
well
in
terms
of
like
looking
at
the
road
map
and
seeing
it
still,
we
actually
do
have
a
couple
initiatives
going
on.
I'm
wondering
if
it's
worth
trying
to
capture
some
of
those
initiatives
as
themes,
as
maybe
milestones.
So
I
know
that
wes
and
dominicus
are
working
on
pkjs.org
and
we've
got
glenn
and
myself
and
a
couple
folks
working
on
docs.
D
So
maybe
we
turn
a
couple
of
those
themes
into
milestones
and
actually
just
make
like
a
kanban
board,
and
then
you
know,
try
and
use
that
I
don't
know
if
we
have
other
themes
to
add,
but
maybe
that
could
be
something
for
us
to
think
about
for
2021.
You
know:
do
we
have
any
other
kind
of
special
projects?
Maybe
one
of
those
is
a
project
evangelism
you
know,
maybe
other.
Is
there
a
way
that
a
couple
folks?
So
basically
we
could
yep.
D
You
know,
break
off
in
a
couple
groups
focus
on
two
or
three
things
at
a
time.
Use
the
milestones
to
track
that
and
the
project
board
to
just
kind
of
you
know,
show
the
work
in
progress
on
the
assembly,
so
to
speak
so
and
then
you
know
certainly
probably
use
more
themes
but
yeah.
D
I
was
just
thinking
about
that
and
you
know
thinking
about
his
question
it
you
know,
there's
it's
a
little
hard
to
grow
kind
of
what
you
don't
even
see
anything
about
pkjs.org
and
that's
kind
of
been
a
big
thing
for
the
past
few
months.
So.
B
Yeah
so
yeah,
maybe
instead
of
just
obliterating
the
project,
we
could
actually
update
it
and
get
it,
get
it
up
to
date
and
make
it
you
know
actually
reflect
reality.
B
Okay,
well,
cool
and
then
so
the
other.
My
other
questions
were
just
kind
of
about
the
pkgs
org
and,
like
you
know,
is
there
some
sort
of
overview
of
the
work
going
on
in
there?
B
I
don't
know
it's
it's
hard
for
me
to
to
jump
into
this
repo
and
then
be
like
okay,
there's
all
this
other
stuff
in
this
package
just
oregon.
I
have
no
idea
what
what
any
of
it
has
to
do
with
any
particular
issue
here
or
like
who's
working
on
what
or
what
needs
help.
So
I
mean
yeah.
D
D
So
that's
something
I
wouldn't
mind
doing
is
maybe
just
trying
to
group
some
of
the
our
ongoing
themes
and
maybe
put
little
descriptions,
and
you
know
who
you
could
reach
out
to
if
you're
interested
in
working
on
said
thing
and
then
try
and
put
the
issues
that
we're
working
on
into
the
current
roadmap
board
and
maybe
move
the
things
that
haven't
had
any
activity
in
a
while,
just
off
the
board,
and
then
that
way
when
you,
if
you
went
to
the
board
you'd,
see
like
oh
wes
and
dominicus
are
dude,
speaking
of
which
he's
got
his
hand
raised
yeah.
D
E
From
one
perspective,
yeah
you're
right,
maybe
we
need
an
overview
of
that,
but
on
the
other
hand,
I'm
rarely
convinced
that
we
need
more
process
in
general
as
a
general
rule,
and
I
think
the
pkgs
it
has
like
what
13
repositories
and
they
kind
of
have
descriptions
there.
The
descriptions
may
not
always
be
great,
but
it's
it's.
It's
not
like.
We
have
a
set
process
and
I
think
that's
a
good
thing.
E
Yeah
and
and-
and
I
think
you
you'll
probably
find
that
in
some
of
the
issues
some
of
the
tools
are
mentioned
there
right,
so
you've
drawn
the
line
between
webby
and
the
original
issue
discussing
sidjim
and
that
kind
of
stuff.
I
think
you'll
find
there's
a
couple
of
smaller
tools
that
wes
has
built
or
has
been
involved
in
and
kicked
off
so
yeah.
B
I
mean
yeah,
I
I
agree
like
I.
I
am
not
asking
for
a
bunch
of
bureaucracy,
but
on
the
other
hand,
it
makes
it
a
little
difficult
to
if
you're,
if
you're
me
like
to
kind
of
jump
in
and
say,
okay
well,
what
what's
going
on
and
what
can
I
help
with?
Because
there
is
not
this
overview,
there's
not
a
high
level
view
of
that
work.
So
that's
in
my
mind,
that's
the
drawback,
but
yeah.
B
I
totally
understand
not
wanting
to
add
process
to
anything
because
that
yeah,
so
I
I
don't
know
you
know
in-
and
it's
also
probably
going
to
be
a
little
more
awkward
to
have
all
the
the
work
happening
in
one
org
and
then
the
project
which
tracks
that
work
in
a
different
word.
Just
because
of
like
the
boundaries
between
you
know
on
github,
where
it's
it's
I
don't
imagine,
you
can
take
an
issue
from
org
a
and
put
it
into
a
card
in
a
project
in
order
b.
B
Maybe
I'm
wrong,
but
yeah
that
that
might
prove
a
challenge.
I'm
not
I'm
not
suggesting
that
you
know
hey.
The
package
chase
needs
its
own
project
board
because
I
I
think
that
will
become
obvious
if
it
if
it
is
an
issue,
but
just
just
from
my
point
of
view,
just
just
as
somebody
who's
who
who's
coming
on
and
wants
to
get
more
involved.
It's
it's
kind
of
hard
to
see
what's
happening.
E
Fair
point
maybe
yeah
just
getting
an
overview
or
just
generating
that
overview
from
the
list.
That's
already
there
on
get
up
just
re-scanning
through
the
projects.
It
is
a
question.
A
B
B
F
Mean
I'm
currently
working
to
try
to
get
meat
into
a
state
where
it's
more
broadly
usable,
so
two
parts
of
that
one
is
ripping
out
the
current
release,
tooling,
that
is
built
in
because
it
it
kind
of
hinders
development,
it's
good
for
automation,
but
it's
not
good
for
development
so
going
through
and
ripping
that
out
or
either
replacing
it
or
creating
a
new
system.
There's
a
different,
a
bit
of
action
that
we
could
use
that
does
relatively
the
same,
but
doesn't
enter
the
development
the
same
way.
F
The
other
thing
is
just
like
adding
a
few
features
like
the
way
west
built.
It
was
largely
like
write
your
own
javascript
to
automate
the
things
and
the
thing
the
big
thing
I'd
like
to
do.
There
is
make
it
so
you
don't
have
to
know
javascript
to
get
the
benefit
of
having
those
things
so
like
being
able
to
just
like
put
in
a
template
some
kind
of
template
kind
of
structure
for
give
me
time,
like
a
default
time
zone
block
of
all
the
major
cities.
F
That
would
be
relevant
to
this
meeting,
which
the
code's
already
written,
like
it's
literally
written
in
the
go
base,
and
just
basically
going
through
and
replacing
that
in
the
the
meeting
template
that
gets
passed
through.
That
kind
of
thing
is
the
next.
The
other
thing
there
that
I'd
like
to
see
us
a
ship
in
that,
because
that
really
lowers
the
barrier
for,
like
you,
know
the
openjs
foundation,
to
use
it
without
having
to
ask
people
to
write
javascript,
which
enables
more
people
to
be
able
to
manage
it
effectively.
B
Yeah,
well,
can
you
just
like
ping
me
in
some
yeah.
G
B
Issue
or
two
sure
I
don't
I
don't
know,
maybe
you
can
even
assign
it
to
me,
I'm
not
sure.
If
I
have
if
I'm
like
in
it
in
package.js,
I
don't
think
I
am,
but
if,
if
I
was,
then
you
could
assign
a
team
so.
A
And
I
guess,
unfortunately,
this
is
probably
our
later
meeting
for
for
beth,
but
she's
she's
been
helping
to
push
forward
the
work
in
webby,
which
is
the
you
know,
one
of
the
things
we're
doing
to
help
module
owners
and
maintainers
basically
check
that
when
they
do
updates
they're
not
breaking
dependencies-
and
I
know
she
has
a
list
of
of
a
few
things
that
could
that
would
be
good
to
help
out
on
that
front.
So
if.
B
You're
reaching
yeah-
I
I
I
reached
out
to
her
at
least
in
an
issue,
but
I
didn't
hear
back
so
I
figured
she's
busy
with
something,
but
I
would
like
to
sync
up
with
her
and
see
like
what.
H
B
H
B
B
So,
no
so
wibby!
Basically
you
know,
I
know
that
we
had
we
being
ibm
at
one
point.
Had
you
know
you,
and
maybe
somebody
else
andrew
or
something
working
on
this,
and
I
was
just
kind
of
wondering
like
what
what
where
you
know
I
just
like.
I
wanted
more
information
about.
B
You
know
what
what
the
intended
use
cases
were
because
like
in
I
don't
have
the
issue
handy,
but
I
was
talking
with
I'm
sorry
is
it
I
don't
even
I'm
on
my
phone,
so
I
can't
even
see
people's
names
dominicus.
Is
that
how
you
pronounce
it
yep?
Okay,
I
was
talking
with
him
about
you,
know.
B
Okay,
so
wibby
does
x,
sikkim
does
y
and
the
the
original
issue
that
I
saw
anyway
was
from
mateo
and
he
said
we
want
something
like
sitcom
and
but
we
want
it
to
be
generalized,
so
other
you
know,
user
land
packages
can
use
it,
but
webby's
doing
something.
At
least
my
man
mine,
you
know
considerably
different
than
what
sitcom
does
and
you
know
it's.
Obviously
it's
a
means
to
an
end.
B
There
are
two
different
ways
of
going
about
solving
the
same
thing,
but
I'm
wondering
like
why
is
it
one
way
and
not
the
other
way,
and
you
know
there's
some
rationale
for
that,
but
I
was
wondering
if
you
know
best
you
or
somebody
who
has
spent
time,
you
know
designing.
This
thing
could
like
explain
more
about
why
it
should
do
what
it
does
instead
of
what
sitcom
does?
Does
that
make
sense.
E
I
can
I
can
give
you
a
very
simple
answer
to
that.
It's
the
the
current
stuff
that
webby
does
is
because
I
wrote
up
a
proposal
to
do
it
that
way,
and
I
wrote
up
the
proposal
to
do
it
that
way,
because
I
I
wanted
to
do
to
work
that
way
right,
there's,
there's
nothing
preventing
functionality
inside
that.
Does
it
the
way
that
you
describe
right,
which
is
a
fair
enough
way
to
do
it
and
yeah.
I
Just
doesn't
scale
right
like
it
requires
a
centralized,
very
expensive
ci.
It
requires
being
able
to
support
the
disparate
ci
setups
that
all
of
the
projects
it's
checking
have
and
like
the
the
intention
is
to
be
able
to
create
a
way
that
sort
of
self-serve
any
maintainer
can
say.
Am
I
about
to
break
the
things
that
depend
on
me
or
not,
and
the
reality
is
that,
like
sichum
includes
a
small
slice
of
the
ecosystem,
most
things
aren't
worthy
of
inclusion
in
it
right,
but
many
more
things
than
that.
I
A
Yeah
and
a
lot
of
it
was
around
the
model
of
it
wasn't
you
know,
I
wouldn't
cast
it
as
just
oh,
hey
one
way
and
you
could
do
it.
Otherwise.
I
think
we
discussed
it
and
we
thought
it
presented
a
better
model
in
terms
of
that
scalability.
In
that
some
some
modules
tests,
you
can't
just
clone
and
run
right.
They
require
some
sort
of
other
other
setup
or
infrastructure,
and
so
by
being
able
to
leverage
these
the
ci
environments
of
the
modules
themselves,
you're
going
to
end
up
with
something
that's
more
functional
and
scalable.
A
The
work
there
so
far
is
is
like
to
do
that
in
an
incremental
way,
so
that,
yes,
you
may
be
running
it
locally
in
our
own
forks
to
start
with,
but
the
end
goal
is
really
to
hopefully
leverage
the
existing
ci's
whenever
possible,
so
that
you
know
you
know
if
you've
got
an
important
module.
Hopefully
your
10
most
significant
dependents
are
going
to
say
sure
I
I
want
you
to
run.
A
B
Okay,
you
know,
that's,
I
think,
that's
the
that
was
the
the
the
point
I
was
I
was
I
was
missing
in
in
so
far
as
like
the
the
you
know,
sitcom,
I
I
it
seems
to
work
pretty
well
like
you
know,
node
is
not
breaking
a
bunch
of
stuff
all
the
time
and
so
like
that
suggests
to
me
that
you
know.
Despite
its
drawbacks,
you
know
it's,
it's
been,
it's
been
good,
it's
definitely.
B
I
Node's
the
platform
knows
the
an
implicit
dependency
of
every
package
and
so
just
swapping
out
the
node,
underneath
it
like
magically
tests
it
on
the
new
node.
But
packages
and
their
dependencies
are
explicit
and
it's
a
little
trickier
to.
I
think,
take
a
dependency
graph
and
just
kind
of
programmatically
swap
out
things
from
inside
it
and.
A
Platform,
we
have
a
robust
ci
with
you
know.
We
have
a
ci
that
can
that
can
support
something
like
sedum,
whereas
I
don't
know
that
every
maintainer
or
a
lot
of
maintainers
won't
necessarily
have
that
same
ci
they're,
relying
on
some
other
ci
and
pulling
in
and
running
the
tests
of
everybody
else
could
actually
become
more
than
like
the
free
limits
will
support
and
that
kind
of
stuff.
B
E
E
A
Right
like
what
we've
got
so
far
is
like
well
at
some
point.
You
know
if
it's
established,
you
might
go
and
say
to
your
dependencies:
hey.
Can
I
run
the
tests
in
your
ci
and
they'll
say
yes
to
start
with,
we
want
to
be
able
to
run
it
all
locally,
so
we
can
make
progress
without
having
to
run
tests
in
other
people's
ci's,
which
I
think
can
lead
to.
What
dominique
has
said
is
the
first
model
where
you
are
running
it.
You
may
still
end
up
with
some
of
the
the
challenges.
A
I
B
A
Bethany
for
some
you
know,
I
think
I.
H
Think
actually,
dominicus
has
labelled
a
project
in
wibby
of
like
mvp,
like
the
first
last
few
issues
to
get
like
the
first
mvp
out
the
way
they're
all
labeled
up.
Yes,.
E
Yeah
we
we,
we
had
the
dock
that
bethany
prepared
originally
in
terms
of
one
there.
I
added
some
checkboxes
there
to
track
the
progress
there.
E
Like
I
I
hear
what
christopher
is
is
asking
for
in
a
tool,
and
he
does
make
a
point
and
if
we
could
build
that
into
webby
as
well
as
all
the
other
things
like
we,
we
now
have
a
configuration
file
format,
it's
very
crude
right
now,
but
the
idea
is
that,
even
even
though
we
discussed
that
you
may
be
opening
pull
requests
in
some
cases,
you
may
need
to
open
pull
request
in
some
cases.
In
some
cases
you
may
need
to
fork.
E
In
some
cases
you
may
be
allowed
to
just
push
a
branch
directly,
so
you
will
have
some
configuration
and
maybe,
in
some
cases
it's
just
enough
for
you
to
do
the
you
know
white
smoke
test
if
you
will
to
just
check
whether
one
version
of
node
passes
very
quickly
without
doing
all
the
other
things
in
a
gym
style.
So
I
mean
yeah,
it's
just
a
matter
of
getting
hands
to
build
that
so.
B
All
right
well,
so
it
is
like
37
minutes
after
the
hour,
so
I'm
sure
there's
other
stuff
on
the
agenda.
So
thank
you.
Okay,
thanks.
A
Right
so
the
next
one
was
number
14
maintainer
needed
query
string.
I
think
owen
was
it.
You
who'd
put
that
on.
C
Quite
a
lot
of
interaction.
Some
people
said
no,
no
we're
good.
D
On
that
particular
issue,
I
did
catch
up.
I
followed
back
to
the
issue
in
query
string.
I
think
tyranny
had
originally
connected
us,
but
it
looks
like
the
team
is
now
the
end.
They
had
at
a
point
in
time
raised
to
maintain
her
needed
flag,
but
I
think
they've
now
found
their
maintainer
and
they're
good
to
go
at
least
from
what
I
was
able
to
gather
in
the
their
thread.
So
on
our
we
should
close
that
one
out.
I
think
we
can
yeah
okay.
D
Yes,
but
make
sure
I
wasn't
missing
anything,
but
I
think
my
interpretation
was
off
but
yeah
sounds
like
we're
in
a
good
spot
to
do
that.
E
That's
why
that's
me,
this
is
actually
something
I
need
help
on
very
much
so
because
I
am
trying
to
solve
a
similar
problem
for
the
purposes
of
webby
as
well,
because
for
wibby
to
actually
be
useful
and
with
the
goal
of
of
trying
to
keep
it
simple
for
maintainers,
it
needs
to
be
able
to
maintain
a
set
of
github
actions.
It
could
be
built
as
a
github
application,
but
I
very
much
dislike
that
model
personally
and
it's
yeah,
so
it
needs
to
be
able
to
maintain
a
set
of
github
actions.
E
So,
as
a
user
of
webby,
you
would
need
to
install
a
couple
of
actions
in
your
repository,
and
that
means
that
you
know
copy-pasting.
These
actions
is
no
go
right.
They're
copy-pasting,
workflows
across
many
repos-
that's
that's
just
wasteful,
but
gear
up
does
not
provide
any
better
primitives
than
actions
which
are
at
step
level
rather
than
the
whole
workflow
level.
So
for
purposes
of
wipi.
E
There
is
a
need
to
have
something
which
allows
you
to
sync
these
things,
or
my
current
thinking
is
that,
for
example,
when
you
npm
install
webby
will
have
maybe
a
post
install
script,
or
maybe
an
explicit
script
that
you
know
maintains
these
actions.
E
You
would
do
something
like
webby
update
actions
and
then
it
would
go
and
update
all
your
workflows
with
the
latest
version
that
it
expects,
rather
not
that
it
expects,
but
just
the
latest
version
that
is
useful
for
maintainers
right,
and
this
gets
back
to
the
issue
of
auto
updating,
node
versions
and
github
actions.
E
With
the
recent
news
from
travis
travis
see,
I
config
travis
is
kind
of
becoming
useless,
but
there's
a
need
for
something
say:
config
github
actions
which
allows
you
to
automatically
drag
in
new
node
versions
into
your
test,
matrix
and
yeah.
E
The
code
am
I
sharing,
I'm
not
sharing
my
so
I
can
quickly
share
my
screen
right,
so
we
can
look
at
at
the
working
example,
the
smallest
one
that
I
could
come
up
with
right
and
this
is,
you
would
have
a
prepare
node
matrix,
which
takes
in
the
ci
configure
actions
which
we
maintain,
which
generates
a
matrix
based
on
your
policy.
So
it
will
take
all
the
lts
versions
which
are
greater
than
10,
and
then
you
use
that
in
a
variable,
but
this
is
very,
very
verbose
right.
I
I
so
just
I've
because
of
travis's
recent
changes.
I've
been
looking
into
migrating,
my
hundreds
of
repos
over
to
github
actions,
and
this
exact
approach
is
what
I
ended
up
settling
on
it's
annoying
and
I
have
to
copy
paste
a
bunch
of
boilerplate,
but
it's
like
in
theory.
It
means
I
don't
have
to
make
any
changes
to
the
local,
the
inline
ci
config
in
the
future.
E
I
What
would
be
amazing
is
if
we
had
a
github
package
that
could
essentially
do
that
like
look
at
a
template
and
stamp
that
onto
the
current
repo
and
then,
like
you
know,
error
if
there's
any
diffs
or
like
you
know,
like
sort
of
a
like
a
way
to,
and
you
know
reusable
beyond,
just
package.js
but
like.
E
D
I
I
I
D
Yeah,
I
wonder
if
they
have
an
issue
tracker,
that
we
could
either
open
an
issue
in
or
you
know,
contribute
to
or
something.
J
Them
there's
community
forums
and
github
support
for
feedback,
but
I
don't
see
any
way
to
to
put
certain
things
like.
We
could
go
poke
the
actions
team
and
I
could
try
to
figure
out
and
find
a
status
update
for
you
on
shared
config
but
yeah.
I
don't.
I
don't
know
off
the
top
of
my
head
and
maybe
tyranny
knows.
D
I
mean,
I
think,
it's
promising
in
the
sense
that
both
you
and
jordan
seem
to
arrive
at
similar
solutions.
So
I
mean
you're
seems
to
be
trending
in
the
right
direction,
with
whatever
limitations
you
have
at
the
moment,
and
you
know
I'd
say:
there's
probably
a
good
reason
to
think
that
github
actions
could
only
get
you
better
going
forward,
whether
that's
three
months
six
months
or
a
year
is,
I
guess,
to
be
determined.
I
Now
sorry,
so
I
eventually
will
want
to
because
I
have
a
lot
of
repos
and
like
it's
a
huge
pain
in
the
butt
like
right
now,
I'm
doing
it
manually
just
because
I
need
my
ci
to
work.
But,
like
you
hear
me,
yeah
essentially.
I
Is
oh,
hey,
jenny,
but
yeah?
Essentially,
what
I'd
like
to
build
is
a
tool
that
does
that
that
just
kind
of
I
run
it
locally
and
it
just
like
does
the
right
thing
where
it
like
makes
a
pr
or
something
on
each
repo
whose
dot
github
folder
doesn't
match.
The
text
you
know
is
missing
stuff
from
the
template
or
something
like
that.
It's
a
little
tricky,
though,
because
some
repos
have
their
own
disparate
config,
like
in
two
of
my
repos.
I
I
run
tests
against
like
ses
a
secure,
ecmascript
thing,
but
I
don't
want
to
do
that
everywhere
else,
but
like
and
also
the
node
version
ranges
that
each
thing
supports
are
a
little
different.
So
I'd
have
to
come
up
with
a
better
story,
but
what
would
be
even
better
is
if
github
just
actions
just
natively
supported
the
ability
to
share
more
than
just
individual
steps
but
to
share
workflows,
because
then
it
would
be
really
easy.
I
would
just
have
a
single
file
that
says
run
these.
I
F
Looks
like
darcy
linked
a
public
road
map
feature
for
that
that
will
be
shipping
in
q2
2021
six
months
of
manual
work
darcy
is
that
is
that
microsoft
q2,
or
is
that
github
cute
or
like
real
world
q2?
Because
I
I
don't
know,
oh.
F
Cool
that
I'm
glad
that
they're
using
real
real
quarters
there,
so
I
mean
the
the
thing
I
think
we
could
do.
There,
though,
is
at
least
at
the
very
least
if
we
want
like
building
a
module
to
that
kind
of,
does
that
work
of
outputting
it?
That
would
still
be
useful
once
that
ships,
you
just
run
it
on
one
repo
rather
than
five
billion
right,
and
so
that
that
might
be
worth
doing.
I
Still
yeah
I
mean
I
can
also
imagine
like
so.
I've
noticed
that
renovate
will
update
like
when
I
have
my.
I
have
some
actions
that
have
the
checkout
action
at
v1
and
renovate
has
started
updating
those
to
v2.
So
there's
some
things
renovate
can
do.
It
seems
like
they're
just
exploring
possibilities
there,
but
I'm
wondering
like
having
a
service
like
that.
Essentially
that
just
kind
of
makes
the
pr's
for
me
would
also
probably
be
nicer
than
me
having
to
run
something
locally
or
put
you
know
anytime.
I
I
have
to
put
stuff
in
actions
on
a
cron.
It
feels
like
a
failure
so
but
yeah,
I
think.
In
other
words,
I
think
github
actions
has
the
potential
to
be
leagues
better
than
every
existing
ci
system,
but
right
now
they're
only
at
like
70
of
of
any
of
them.
So
I'm
hoping
that
you
know
they'll
staff
that
up
and
keep
improving.
E
All
right,
okay,
would
you
should
we
open
a
new
issue
and
post
some
use
cases
there,
because
I
know
that
I
will
have
this
problem.
I
don't
know
when,
but
because
github
enterprise
lags
behind
kit
of
public
by
six
to
12
months.
I
will
need
this
solved
at
some
point.
So
if
there's
use
cases
to
consider,
then
maybe
another.
E
E
I'll
add
an
issue
and
start
collecting
some
use
cases
and
then
I'll
bring
some
people
I'll
think
mateo,
because
I'm
sure
he'll
run
into
this
pretty
soon
and
I'll
pick
maybe
bring
somebody
else
in
their
form
and
see
if
there's
anybody
who's
actually
started
building
anything
like
that
and
like
I'm
sure
that
somebody
has
to
have
built
something
for
that
by
now.
It's
just
that.
It's
really
hard
to
find.
I
And
something
else
I
did
dominique
is,
I
don't
know
if
your
your
matrix
action
code
does
this
but
mine.
Actually,
I
made
it
fetch
the
json
file
from
nodejs.org,
so
the
versions
are
always
up
to
date,
so
the
instant,
a
new
node
version
is
released.
It's
automatically
included
in
the
appropriate.
Oh.
E
I
F
One
one
other
thing
that
I
I've
thought
a
lot
about
is
like:
if
do
we
want
to
maintain
some
like
kind
of
like
recommended
ci
like
layout,
for
node
or
from
node,
rather
not
for
node,
but
from
node?
So
like
here's,
what
we
recommend
you
use
and
people
can
like
we
can
toss,
we
can
give
them
a
boilerplate
template
that
will
set
them
up
for
that,
including
auto
updating
or
whatever.
F
C
A
Session
because
we've
only
got
10
minutes
left,
so
I'm
thinking
we've
got
the
next
step
for
this
one.
The
next
issue
suggests
a
list
of
modules
to
get
to
get
support
info
into.
I
still
have
noted
an
ap
on
list.
The
red
hat
team
is
working
to
get
the
support
info
added
to
a
number
of
node
shift
modules,
but
if
other
people
could
help
get
it
into
modules
or
or
add
in
that's
what
that
issue
is
for.
A
If
people
can
think
about
that
and
then
the
suggested
during
vulnerability
by
the
package
maintainer,
I
know
wes
and
darcy
were
working
on
a
proposal
for
that
front,
but
I
haven't
been
able
to
catch
up
with
wes
lately.
I
don't
know
if
darcy
you've
talked
to
him
lately.
J
Unfortunately,
not
the
last
month
has
been
a
little
bit
hectic.
We've
shipped
about
12
releases
of
the
mpm
client,
so.
A
E
Just
a
very
quick
yep
I
have
I
just
yeah.
I
have
two
pull
requests
that
I
wanted,
but
I
have
two
approvals
on
one
of
those.
So
that's
four
two
two.
If
I
can
get
some
more
please
and
then
add
the
larger
one,
the
dependency
management
one
I
messed
up
so
now
it's
open
from
unknown
repository,
because
I
deleted
my
fork
and
now
I
can't
update
it,
but
I
do
have
another
one
which
follows
up
the
issues
that
jordan
has.
E
I
So
yeah
just
awesome,
tierney
or
darcy
any
or
roy.
I
suppose
any
chance
that
we
could
sneak
in
somewhere
into
github's
roadmap
away
to
prevent
people
from
unknowingly
screwing
up
open
prs
by
deleting
a
fork.
People
do
that
on
pr's
they've
made
on
my
repos
all
the
time
and
it
really
sucks
because
there's
no
recovering
it
and
I
can
never
rebase
that
pdr
again.
F
I
can
I
can
pass
that
feedback
along.
I
I
don't
think
I
can
get
it
prioritized,
but
I
can
pass
it
along.
I
I'm
more
than
happy
to
schedule
conversation
time
to
implore
for
it.
I
just
don't
know
who
to
talk
to
so.
C
A
A
So
we
I'm
going
to
stop
the
live
stream
now
and
we'll
move
to
a
private
session.
If
that's
okay,.